Conceptualisation

The way in which financial exclusion is measured depends on how it is conceptually defined. This chapter follows the World Bank’s definition of financial exclusion as a lack of access to useful and affordable financial products and services that meet the individual’s needs for transactions and payments, savings, credit and insurance (World Bank, 2017).

According to this definition, one pillar of financial exclusion is the ability of individuals to access services that allow them to purchase goods and services and store wealth. Macroeconomic studies usually proxy this indicator with the size of the ‘banked’ population, or the proportion of people with access to a transaction account (Cámara & Tuesta, 2014; Chakravarty & Pal, 2013; Massara & Mialou, 2014; Park & Mercado Jr, 2015; Sarma, 2008). A microeconomic counterpart to this measure could simply be measured with variables that determine whether a household or individual has access to a checking account.

Access to savings is another pillar of financial exclusion. Intuitively, households forego current consumption to increase future consumption to maximise inter-temporal utility. Macroeconomic studies traditionally incorporate savings into financial exclusion

2014). A microeconomic approach to measuring savings is similar. Access to savings can be determined using variables that capture whether households have access to bank term deposits, stocks, funds, bonds and other financial products. Bodie, Merton and Cleeton (2009) argued that these financial instruments can be grouped into three categories: debt, including term deposits and bonds; equity, including stocks; and derivatives, including futures, forward contracts, swaps, and other financial instruments. While savings essentially allow households to smooth consumption through time by foregoing present consumption for future consumption, credit allows households to forego future consumption for present consumption. Macroeconomic studies have typically included measures, such as the proportion of people who receive credit, as a measure of this dimension (Cámara & Tuesta, 2014; Sarma, 2008). A microeconomic counterpart can similarly be based on whether a household or individual has a loan or credit card.

Finally, households also use financial services for insurance to help build resilience against covariate and idiosyncratic shocks (Bodie et al., 2009). This need can be met by purchasing various types of insurance products, including life insurance, health insurance and property insurance. Macroeconomic studies do not usually include insurance controls. The microeconomic approach can simply capture this information with a question regarding whether households have access to insurance services.

4.3

Methodology

To compute the MFEI, this chapter adopts a strategy similar to the computation of the MPI by Alkire and Santos (2010, 2013), Alkire and Foster (2010), Dotter and Klasen (2017) and Jāhāna (2015). The methodology of the MPI is borrowed because it is the most prominent new development index that focuses on household-level and microeconomic indicators. The MPI has been estimated for more than 100 economies since 2010, with the results published annually alongside the HDI in the United Nations Development Programme’s Human Development Report (United Nations Development Programme, 2010).

The MPI is calculated as η x α, where η is the headcount or the percentage of people who are identified as multidimensionally poor and α is the average intensity of poverty among the poor. The headcount measure, η, is similar to the widely used headcount

poverty ratio (Foster, Greer, & Thorbecke, 1984). It has the advantage of being easy to communicate and allows for comparisons with existing poverty measures. The inclusion of the average intensity, α, ensures that the MPI simultaneously concerns itself with identifying the incidence of poverty, as well as the depth of the deprivation being experienced. The MPI includes information on 10 indicators of wellbeing that are grouped into three equally weighted dimensions: health, education and a non-monetary standard of living (United Nations Development Programme, 2010).

Consequently, to compute the MFEI, this chapter assigns each household a deprivation score 𝑐 using the indicators described in the previous section. A cut-off of 50 per cent, which is equivalent to half of the weighted indicators, is used to distinguish between excluded and included; therefore, if 𝑐 is smaller than 50 per cent, the household is included, and vice versa.23 In the latter case, this chapter further distinguishes two cases: moderately excluded households, in which the deprivation score is greater than 50 per cent but less than or equal to 75 per cent; and severely excluded, in which the score is greater than 75 per cent.

Similarly, the headcount ratio, 𝐻, measures the proportion of people who are financially excluded in the population as:

𝐻 =𝑞_𝑛, (4.1)

where 𝑞 is the number of excluded people and 𝑛 is the total population. The intensity of financial exclusion, 𝐴, reflects the proportion of the weighted component indicators in which, on average, people are deprived. A is measured as:

𝐴 =∑ 𝑐𝑞𝑖 𝑖

𝑞 (4.2)

where 𝑐_𝑖 is the deprivation score that the 𝑖th excluded individual experiences.

As with the MPI, the value of the financial exclusion index is therefore the product of the headcount ratio and the intensity of financial exclusion, which means that the MFEI is given by:

23 _{A 50 per cent cut-off is chosen for consistency between this measure and the MPI. Basically, there are}

four dimensions of financial inclusion/exclusion: transactions and payments, savings, credit and insurance. Each dimension is treated equally because they each form an equal component of a household’s basic

𝑀𝐹𝐸𝐼 = 𝐻 × 𝐴 (4.3) To assess which components are driving financial exclusion, the contribution of dimension 𝑗 to multidimensional financial exclusion can be calculated as:

𝑐𝑜𝑛𝑡𝑟𝑖𝑏_𝑗 =∑ 𝑐𝑞1 𝑖𝑗

𝑛 /𝑀𝐹𝐸𝐼. (4.4)

Calculating the contribution of each dimension to multidimensional financial exclusion provides information that can be useful for revealing a country’s configuration of deprivations, and it can help with policy targeting. Table 4.1 summarises each indicator and its weight. Following Alkire and Foster (2010), this chapter uses equal weights for each indicator.

Table 4.1: Dimensions, indicators, deprivation thresholds and weights for MFEI

Dimension (weight) Indicator (weight) Deprived if…

Transaction Checking account

Household has no checking account

(1/4) (1/4)

Saving (1/4)

Debt

Household has no term deposits or bonds (1/8)

Equity

Household has no stock trading account (1/8)

Credit (1/4)

Loan

Household has no mortgage or car loan (1/8)

Credit card

Household has no credit card (1/8)

Insurance (1/4)

Commercial insurance

Household has no commercial insurance (1/4)